gitmylo / bark-data-gen

Create training data for training a voice cloner for bark text to speech.
MIT License
44 stars 10 forks source link

Is it possible to use literature and corresponding wav file from a different language to train my own checkpoints? #2

Open randomradio opened 1 year ago

randomradio commented 1 year ago

I have a collection of audiobooks in Cantonese along with all the text.

I assume following the training process mention in your other repo would provider a model that better fitting for Cantonese?